lepton
1 W
Nice work! Major kudos. Is the motor as loud in person as it seems in the video?
high power/high speed BLDC controller build
- Thread starter dirty_d
- Start date
WOW, awesome work there 
. But I wouldn't worry about the temperature, 4.5A @ 38V is 171W of heat going into that motor, so it should get pretty warm. However, if you want to use this controller with 6 fet's on an EV, you'll need to add some serious current limiting to the software. I think if you ganged up about 6-8 FET's per leg (36-48 total) it would be one awesome setup with tire smoking power. My scooter is running a Jeti 200A esc....but sensorless really sucks during high acceleration and low speeds. Also, the throttle on the Jeti is a straight PWM control which is no good on an EV. It would be much better to have throttle position = peak torque output. That will give a LOT smoother ride, much better range, but still provide the HUGE acceleration this motor is capable of.
. But I wouldn't worry about the temperature, 4.5A @ 38V is 171W of heat going into that motor, so it should get pretty warm. However, if you want to use this controller with 6 fet's on an EV, you'll need to add some serious current limiting to the software. I think if you ganged up about 6-8 FET's per leg (36-48 total) it would be one awesome setup with tire smoking power. My scooter is running a Jeti 200A esc....but sensorless really sucks during high acceleration and low speeds. Also, the throttle on the Jeti is a straight PWM control which is no good on an EV. It would be much better to have throttle position = peak torque output. That will give a LOT smoother ride, much better range, but still provide the HUGE acceleration this motor is capable of.
dirty_d
10 kW
yea its pretty loud, it works but i may have the hall sensors between the wrong poles still. i also just destroyed at least 1 fet for some unknown reason, this is getting frustrating.
dirty_d
10 kW
CNCAddict, yea im going to use 12 fets in the final design, this is just for testing, i dont wanna risk that many fets at one time. im finding out these small fets are quite the pain in the ass to work with when you want to controll 100+ amps. it would be nice to get some of those really big fets with the screw terminals, i think they are really expensive though.
wrobinson0413
1 kW
.
dirty_d
10 kW
after i disconnected the shorted fet and blew another and a gate driver i stomped the shit out of the controller boards in a fit of rage and bloodied up my sandal clad foot 
. i really do have a bad temper.
. i really do have a bad temper.
johnrobholmes
10 MW
Which way were the hall sensors oriented?
Yeah...I feel your pain. The funny part is that you have gotten the motor to actually spin. That's a lot better than most of us do. Anyway, I think you're referring to IBGT's when you talk about the screw on terminals. You can find them for a lot less on ebay.
http://search.ebay.com/search/search.dll?from=R40&_trksid=m37.l1313&satitle=igbt&category0=
http://search.ebay.com/search/search.dll?from=R40&_trksid=m37.l1313&satitle=igbt&category0=
dirty_d
10 kW
john, they are all facing outwards for 120 degree signals.
its just frustrating because the power components are pretty expensive.
its just frustrating because the power components are pretty expensive.
wrobinson0413
1 kW
.
johnrobholmes
10 MW
I am going to duplicate your experiment with hobby parts now. See if I can get an outrunner to spin up on a sensored hobby ESC.
Could you recommend a sensor to start with?
Could you recommend a sensor to start with?
dirty_d
10 kW
john, i used these http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=480-1999-ND, they are open collector, so you need a pull up resistor on the output pin.
dirty_d
10 kW
this is the driver board i made for it, if anyone wants the schematic or pcb layout files just ask.
dirty_d
10 kW
johnrobholmes
10 MW
Got some sensors on order today, here is to hoping I can duplicate this with one of my motors! If so, I think we may have a HUGE opening for sensor based outrunners that could work with current bike controllers assuming the commutation speed is up to the task.
dirty_d
10 kW
yea no reason they wouldnt work as long as the controller can keep up.
johnrobholmes said:
It should work fine. I've seen little 6 pole inrunners with halls that go over 80,000 rpm.
johnrobholmes
10 MW
I picked up some smaller sensors at the same time as the ones you recommended -- http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=AN48841B-NLCT-ND
The smaller ones will fit in the space better, I just hope they will work like a hall sensor should. This is the pinout that is standard for RC style controllers
It must use a six position JST ZH connector model number ZHR-6 or equivalent
connector with 6 JST part number SZH-002T-PO.5 26-28 awg contacts or equivalent.
Wire sequence must be as follows:
Pin #1- Black wire-ground potential
Pin #2- Orange wire-phase C
Pin #3- White wire-phase B
Pin #4- Green wire-phase A
Pin #5- Blue wire-temp control, 10 k Thermistor referenced to ground potential
Pin #6- Red wire-+ 5.0 volts =/- 10%
So my best guess is that all of the sensor pin 3's will tie together for ground, all pin 4's will tie together for + 5v, and the pin 5's will be tied into the respective phase pins for the controller-- Pins 2 through 4. Seem right on track? This will be very useful, and this thread has already increased my understanding of sensored motor construction greatly.
The smaller ones will fit in the space better, I just hope they will work like a hall sensor should. This is the pinout that is standard for RC style controllers
It must use a six position JST ZH connector model number ZHR-6 or equivalent
connector with 6 JST part number SZH-002T-PO.5 26-28 awg contacts or equivalent.
Wire sequence must be as follows:
Pin #1- Black wire-ground potential
Pin #2- Orange wire-phase C
Pin #3- White wire-phase B
Pin #4- Green wire-phase A
Pin #5- Blue wire-temp control, 10 k Thermistor referenced to ground potential
Pin #6- Red wire-+ 5.0 volts =/- 10%
So my best guess is that all of the sensor pin 3's will tie together for ground, all pin 4's will tie together for + 5v, and the pin 5's will be tied into the respective phase pins for the controller-- Pins 2 through 4. Seem right on track? This will be very useful, and this thread has already increased my understanding of sensored motor construction greatly.
dirty_d
10 kW
im kinda confused, what controller are you using?
johnrobholmes
10 MW
I will be using this one: http://www.teamtekin.com/rs.html for initial testing. This will be an RC project before I try the same deal on a bike.
The main Tekin engineer/pres said I wouldn't need a pulldown resistor. Don't know if that is right or not.
The main Tekin engineer/pres said I wouldn't need a pulldown resistor. Don't know if that is right or not.
dirty_d
10 kW
it most likely already has the pullup resistors in the controller.
johnrobholmes
10 MW
dirty_d
10 kW
holy crap, i wouldnt wanna work with those.
johnrobholmes
10 MW
And I didn't get anywhere with them. I was hoping for a bit more leg on em. They just stuck to the soldering iron, and it really made me realize how shaky my hands are (and they are not really)! Now I just gotta get the bigger ones crammed into the motor.
Similar threads
